|Publication number||US4688089 A|
|Application number||US 06/886,178|
|Publication date||Aug 18, 1987|
|Filing date||Jul 15, 1986|
|Priority date||Jul 25, 1985|
|Also published as||DE3526596A1, DE3526596C2|
|Publication number||06886178, 886178, US 4688089 A, US 4688089A, US-A-4688089, US4688089 A, US4688089A|
|Original Assignee||Standard Elektrik Lorenz Ag|
|Export Citation||BiBTeX, EndNote, RefMan|
|Patent Citations (8), Referenced by (5), Classifications (5), Legal Events (6)|
|External Links: USPTO, USPTO Assignment, Espacenet|
The invention pertains to a circuit arrangement for detecting horizontally directed motion of an edge in a television picture.
In conventional television-picture reproduction systems, successive fields are displayed on a screen using interlaced scanning. For so-called frame reproduction, two temporally adjacent fields one of which was delayed by the transmission time of a field are simultaneously presented on the screen by interleaving the lines of the two fields. FIG. 1a shows schematically three successive fields 8, 9 and 10 with scanning lines 1 to 7. FIG. 1b shows the frames 11 and 12 formed from those fields. The fields have a vertical edge 13 at which lines 1 to 7 change from a dark picture area 14 to a bright picture area 15. In the course of the scene, this edge 13 moves in the horizontal direction toward the right margin 16 of the picture, so that the edge 13 is in the location x1 of the first field at the time t1, in the location x2 of the second field at the time t2, and in the location x3 of the third field at the time t3. Since the locations x1, x2 and x3 in the three fields 8, 9 and 10 are shifted relative to each other, in the first frame 11 formed from the two fields 8 and 9, the edge-forming transitions at x1 are shifted relative to the edge-forming transitions at x2. In the display of a frame, the moving edges 13 of the fields thus become a blurred bar 13' whose width is equal to the distance between x2 and x1 and which, on close examination, has a serrate structure.
To reduce or eliminate such picture unsharpness caused by the comblike serrate structure at a moving edge, it is first necessary to detect the places of such motion blur in the television frame and to produce a position-dependent detection signal. Such motion detection can also be used for other frame-signal-processing purposes, such as frame reproduction at increased frame frequency, frame reproduction with noise elimination systems, or frame reproduction with cross-color reduction in the PAL standard.
German Patent Application No. P34 44 836 discloses a detection circuit which requires three field delay circuits of which at least two are connected in series. By a subtracter, the Line signals of every first and third field are locally compared, and if the signal contents disagree, an output signal is produced which, after formation of the absolute value, amplitude weighting, and passage through a further delay circuit, is outputted as a detection signal. That detection circuit requires a large amount of circuitry since each field delay circuit for temporarily storing the information of a field represents a memory circuit with considerable memory location requirements.
One object of the invention is to provide a detection circuit arrangement which only needs a relatively small amount of storage.
A motion detection circuit in accordance with the invention requires only a single field delay circuit, so that it needs a considerably smaller amount of storage than conventional motion detection circuits. In addition, it has been found that the motion detection in conventional motion detector circuits fails if particular speeds of motion of periodic picture structures occur during frame reproduction, particularly if the picture structure moves by just one cycle during two field periods. Such a failure is avoided in the motion detection circuit according to the invention by detecting transitions in the vertical rather than the horizontal direction, namely by the vertically oriented detection of periodically recurring comblike picture structures having a defined vertical local frequency.
To detect horizontally moving portions of a TV frame formed from two adjacent fields, the vertical local frequency at a vertical edge of the moving portion of the picture is measured with a narrow-band-pass filter circuit which is tuned to the local frequency of the comblike structure at a vertical edge of staggered lines of the two fields of the frame. When such an edge structure is detected, the circuit delivers a detection signal.
The narrow-band-pass filter circuit used in the motion detector circuit according to the invention can be implemented using linear signal-processing principles, e.g., as a correlation filter, or non-linear techniques. Further in accordance with the invention an improved band-pass filter circuit used requires little circuitry and has a very narrow bandwidth, a very short transient time and delivers a relatively large output signal. The narrow-band-pass filter circuit is designed to produce a usable output signal already after few (three or four) input pulses.
The invention will be better understood from a reading of the following detailed description in conjunction with the drawing in which:
FIGS. 1a and 1b are schematic representations of adjacent fields and of frames formed therefrom;
FIG. 2 is a block diagram of a motion detector circuit arrangement, and
FIGS. 3 and 4 are more detailed block diagrams of motion detector circuit arrangements.
The block diagram of FIG. 2 shows the basic configuration of a motion detector circuit 18 connected to a video signal line 17 of a television set (not shown in detail). The circuit arrangement 18 includes a frame-signal circuit 19 and a narrow-band-pass filter circuit 20. The pass frequency fo of the band-pass filter circuit 20 is tuned to the so-called vertical local frequency of the comblike serrate structure of the edge stripe 13' resulting from the movement of a vertical edge between the places x1 and x2 of the frame 11 of FIG. 1b. Consequently, the pass frequency of the narrow-band-pass filter circuit is equal to the number of line pairs per frame which are formed from the corresponding lines of the two fields of the frame, multiplied by the number of frames per second. A line pair of the frame 11 of FIG. 1b is formed, for example, by the two adjacent lines 1 and 2, with line 1 coming from the first field 8 of FIG. 1a, and line 2 being the corresponding line of the second field 9. The structure of the edge stripe 13' shown in the frame 11 of FIG. 1b is typical of a horizontally moving picture area with a vertical edge portion. Therefore, an output signal SB of the band-pass filter circuit 20, which is tuned to this characteristic frequency fo, indicates unambiguously any horizontal motion of the detected picture area. This output signal is applied to a video-signal processor 21 which creates a frame (e.g., 11 in FIG. 1b) from the video signals on the video-signal line 17, which were produced by interlaced scanning, and video signals of two successive fields, which were delayed by one field period in the frame-signal circuit 19 and applied over a line 22. This frame is displayed on the screen 23 of a picture tube 24.
In the embodiment of a motion detector circuit arrangement shown in FIG. 3, the frame-signal circuit 19 consists of a field delay circuit 25 having its input 27 connected to the video-signal line 17, and a line delay circuit 26 connected to the output of the circuit 25. The field delay circuit delays the picture signals appearing at its input 27 by one field period, and the line delay circuits introduces a delay equal to the period of one line of this field. Connected to the input 27 and the output 28 of the field delay circuit and to the output 29 of the line delay circuit 26 are two adders 30 and 31 each having a positive input and a negative input, which are part of a computing circuit 32 forming the narrow-band-pass filter circuit. The output of the first adder 30, which forms a first difference value S21 from the input signal S1 and the output signal S2, is coupled to an output channel 33 which contains an absolute-value device 34 and a reciprocal-value device 35 and leads to a first input 36 of a multiplier 37. The output of the second adder 31, which forms a second difference value S31 from the input signal S1 of the field delay circuit 25 and the output signal S3 of the line delay circuit 26, is coupled through a second absolute-value device 39 to a second input 38 of the multiplier 37. The multiplication of the absolute value of the reciprocal of the first difference value S21 by the absolute value of the second difference value S31 gives a signal S (x) at the output of the multiplier 37 which is dependent on the location x in the scanning line according to the following function:
where fo is the frequency to which the computing circuit 32 forming the band-pass filter circuit is tuned, represented as the vertical local frequency, e.g., fo=312.5 cycles per picture height, and f (x) is the vertical local frequency of the picture pattern in the place x of the detected frame. This computing function has the advantage of being independent of the location x and of the modulation depth of the input signal, and of rapidly tending to infinity, which is equivalent to a narrow bandwidth.
Because of the limited accuracy of computation and the quantization of the signal in a digital system, however, this independence of the place of detection and of the limit value "infinity" is not actually attainable. Therefore, the output of the multiplier 37 is followed by a weighting circuit 40 which subjects the output signal S (x) of the multiplier to a nonlinear weighting function 41, so that quite narrow-band detection is achieved for the vertical local frequency fo.
The output signal S2 of the field delay circuit 25 is also applied to the first input 42 of the video-signal processor 21 for displaying a frame on the screen 23 of the picture tube 24 and, through a line interpolation circuit 43, to a first input 44 of a soft-changeover circuit 45. The second input 46 of the soft-changeover circuit is connected to the video-line 17, while the output of the soft-changeover circuit is coupled to the second input 47 of the video-signal processor 21. In the absence of an output signal SB from the weighting circuit 40, whose output is connected to a control input 48 of the soft-changeover circuit, essentially the video signal S1 is applied through the soft-changeover circuit to the second input 47 of the video-signal processor. In the presence of an output signal SB from the weighting circuit 40, which indicates a moving section of the picture, an increasing share of the output signal S2 of the field delay circuit 25, which signal was modified by the circuit 43, is applied to the second input 47 of the video-signal processor in accordance with the amplitude of the signal from the weighting circuit. By this, any motion blur of the vertical edges of horizontally moving portions of the picture is largely prevented.
The motion detector circuit shown in FIG. 4 differs from the motion detector circuit 18 of FIG. 3 essentially in that a second line delay circuit 49 is connected to the input 27 of the field delay circuit 25, and that the computing circuit 32 includes a third adder 50 with a positive input and a negative input. This adder 50 forms a third differences value S34 from the video signals S3 at the output 29 of the first line delay circuit 26 and the video signals S4 at the output 51 of the second line delay circuit 49. From the difference value S21 of the first adder 30 and the difference value S21 of the first adder 30 and the difference value S34 of the third adder 50, a fourth adder 52 in the output channel 33 forms a sum value S21', which takes the place of the difference value S21 of the first adder 30 in the computing process. This increases the reliability of detection of motion in a picture considerably and results in a more definite output signal S (x) of the multiplier 37.
|Cited Patent||Filing date||Publication date||Applicant||Title|
|US3890462 *||Apr 17, 1974||Jun 17, 1975||Bell Telephone Labor Inc||Speed and direction indicator for video systems|
|US4090221 *||Mar 13, 1972||May 16, 1978||Bell Telephone Laboratories, Incorporated||Apparatus for improving video signal-to-noise ratio|
|US4232338 *||Jun 8, 1979||Nov 4, 1980||Bell Telephone Laboratories, Incorporated||Method and apparatus for video signal encoding with motion compensation|
|US4296436 *||Aug 21, 1979||Oct 20, 1981||Hitachi, Ltd.||Noise reducing system|
|US4307420 *||Jun 2, 1980||Dec 22, 1981||Nippon Hoso Kyokai||Motion-compensated interframe coding system|
|US4539594 *||Feb 23, 1984||Sep 3, 1985||Robert Bosch Gmbh||System for reducing noise in a television signal|
|US4626891 *||Jul 23, 1984||Dec 2, 1986||Hitachi, Ltd.||Motion detecting circuit utilizing inter-frame difference signals of successive fields|
|US4636862 *||Feb 7, 1985||Jan 13, 1987||Kokusai Denshin Denwa Kabushiki Kaisha||System for detecting vector of motion of moving objects on picture|
|Citing Patent||Filing date||Publication date||Applicant||Title|
|US4837624 *||Apr 30, 1987||Jun 6, 1989||Robert Bosch Gmbh||Recognition and compensation of errors in a digital video signal|
|US4894716 *||Apr 20, 1989||Jan 16, 1990||Burle Technologies, Inc.||T.V. motion detector with false alarm immunity|
|US5027201 *||Dec 21, 1989||Jun 25, 1991||Rca Licensing Corporation||Motion detection apparatus as for an interlace to non-interlace scan converter|
|US5668609 *||Sep 8, 1994||Sep 16, 1997||Tektronix, Inc.||Motion detector and key signal interpolator using same|
|US20050041101 *||Aug 24, 2002||Feb 24, 2005||Markus Weber||Method and device for identifying motion in an image|
|U.S. Classification||348/700, 348/E05.065|
|Jul 15, 1986||AS||Assignment|
Owner name: STANDARD ELECTRIK LORENZ AG, 7 STUTTGART 30, (FEUE
Free format text: ASSIGNMENT OF ASSIGNORS INTEREST.;ASSIGNOR:UHLENKAMP, DIRK;REEL/FRAME:004578/0636
Effective date: 19860707
|Oct 4, 1988||AS||Assignment|
Owner name: NOKIA GRAETZ GESELLSCHAFT MIT BESCHRANKTER HAFTUNG
Free format text: ASSIGNMENT OF ASSIGNORS INTEREST.;ASSIGNOR:STANDARD ELEKTRIK LORENZ AG;REEL/FRAME:004998/0819
Effective date: 19880913
Owner name: NOKIA GRAETZ GESELLSCHAFT MIT BESCHRANKTER HAFTUNG
Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNOR:STANDARD ELEKTRIK LORENZ AG;REEL/FRAME:004998/0819
Effective date: 19880913
|Nov 7, 1990||FPAY||Fee payment|
Year of fee payment: 4
|Nov 4, 1994||AS||Assignment|
Owner name: NOKIA (DEUTSCHLAND) GMBH, GERMANY
Free format text: CHANGE OF NAME;ASSIGNOR:NOKIA GRAETZ GESELLSCHAFT MIT BESCHRANKTER HAFTUNG;REEL/FRAME:007188/0959
Effective date: 19920706
|Jan 30, 1995||FPAY||Fee payment|
Year of fee payment: 8
|Feb 9, 1999||FPAY||Fee payment|
Year of fee payment: 12